JOURNAL OF MAGNETIC RESONANCE IMAGING 25:453– 465 (2007)

Review Article

Head and Neck Imaging: The Role of CT and MRI Franz J. Wippold II, MD1–3* to an accurate diagnosis of many benign processes as well. A complete study of head and neck imaging is beyond the scope of this article; however, a review of some of the applications of CT and MRI may be helpful.

High-resolution computed tomography (CT) and magnetic resonance imaging (MRI) have become indispensable tools for the evaluation of conditions involving the head and neck. Complex anatomic structures and regions, such as the orbit, skull base, paranasal sinuses, deep spaces of the neck, larynx, and lymph nodes, require that the radiologist be familiar with the imaging modalities available and their appropriate applications. The purpose of this article is to review the techniques of CT and MRI and the roles they play in clinical practice, including head and neck disorders.

KEY CONCEPTS

Key Words: magnetic imaging resonance; CT; head and neck imaging; imaging applications; imaging techniques J. Magn. Reson. Imaging 2007;25:453– 465. © 2007 Wiley-Liss, Inc.

THE RAPIDLY EXPANDING INTEREST in the neck has been largely fueled by the tremendous technical development and availability of cross-sectional imaging, and by the appreciation of these advancements by our clinical colleagues. High-resolution rapid computed tomography (CT) and magnetic resonance imaging (MRI) have long since proven themselves sensitive and reliable in appropriate applications (1,2). Indeed, imaging has become an indispensable tool in the characterization and staging of conditions involving the head and neck, and clinicians have come to incorporate imaging data with physical examinations and endoscopy. CT and MRI not only provide essential information about the deep extension of clinically detected masses, they can also delineate additional clinically unsuspected lesions. The excellent tissue characterization of MR scans can lead

Several key concepts must be applied when considering cross-sectional imaging in the evaluation of head and neck lesions. First of all, imaging is a tool that supplements and complements the physical examination. Cross-sectional imaging is not a “stand-alone” procedure. Second, CT and MRI emphasize anatomy and the changes in anatomy that occur with pathology. Therefore, knowledge of head and neck anatomy is crucial for the accurate interpretation of images. Third, CT and MRI complement each other. Certain processes are better studied with one method than the other, and the various applications, strengths, and weaknesses of each modality should be carefully considered. Fourth, “context is king.” The interpretation of imaging studies should take into account the patient’s history, physical findings, comorbidities, and previous procedures that may influence the structures visualized. Comparison films and previous imaging reports are also extremely useful and enable the imager to understand the clinical issues that prompted the scan request. Of course, old reports may be inaccurate and the radiologist should carefully verify the source of information. TECHNIQUES CT

1 Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, Missouri, USA. 2 Department of Radiology, Barnes-Jewish Hospital South, St. Louis, Missouri, USA. 3 Department of Radiology/Nuclear Medicine, F. Edward He´bert School of Medicine, Uniformed Services University of the Health Sciences, Bethesda, Maryland, USA. Presented at the 13th Annual Meeting of ISMRM, Miami Beach, FL, USA, 2005. The opinions and assertions contained herein are the private views of the author and are not to be construed as official or as reflecting the views of the United States Department of Defense. *Address reprint requests to: F.J.W., Neuroradiology Section, Mallinckrodt Institute of Radiology, Washington University Medical Center, 510 South Kingshighway Blvd., St. Louis, MO 63110. E-mail: [email protected] Received February 21, 2006; Accepted September 28, 2006. DOI 10.1002/jmri.20838 Published online 5 February 2007 in Wiley InterScience (www. interscience.wiley.com).

© 2007 Wiley-Liss, Inc.

Standard Spiral CT CT scanning of the head and neck should be tailored for the anatomic region under consideration. A digital lateral scout radiograph may assist in planning the scan ranges. Spiral (helical) CT scanning is rapidly replacing conventional dynamic CT (slice-by-slice acquisition) in most medical centers (3). Spiral CT, which involves the continuous rotation of the x-ray tube and detector row as the patient translates through the scanning gantry, permits rapid scanning of large volumes of tissues during quiet respiration, usually reduces the amount of needed intravenous contrast material, eliminates much of the motion artifact seen with slower scans, and allows multiplanar and three-dimensional (3D) reconstructions (4 – 6). Moreover, multirow detector technology, which entails rotating the x-ray tube while

453

454

simultaneously employing multiple parallel detector arrays rather than a single detector row, has further advanced spiral CT by reducing scan time and greatly increasing anatomic coverage. Debilitated, elderly, or arthritic patients may require scanning in the supine position; however, reformatted coronal images are usually easily derived from the axially acquired data. With the widespread application of spiral technology and filmless picture archiving and communications systems (PACS), reformatted coronal and sagittal images may be obtained at the workstation, obviating the need to scan the patient in multiple planes. General surveys typically cover the base of the skull to the clavicles with 4- or 5-mm-thick slices. The use of intravenous contrast is recommended. For special regions of interest (ROIs), patients may be scanned with additionally acquired 2-mm slices and a higher zoom factor, or, with newer scanners, the same information can be obtained using reconstructed spiral data. In patients without significant dental restorations, scanning can usually be accomplished in a single range with slices parallel to the laryngeal ventricle. In patients with numerous dental restorations, scan slices should be angled to avoid the teeth.

Wippold

radation due to laryngeal motion or a poor contrast bolus. Three-dimensional (3D) CT techniques may be helpful for radiotherapy planning. This technique can be used to design radiotherapy beam trajectories and thus limit extraneous collateral radiation to other organs, such as the salivary glands (8). CT is also useful for guiding percutaneous biopsies (9). PET, which is regarded as a functional modality, detects lesions by imaging the uptake of the intravenously injected radioactive glucose analog 2-[18F]-fluoro-2-deoxy-D-glucose (FDG) in metabolically active tumors. Combined PET/CT superimposes the functional PET data on nearly simultaneously acquired anatomic CT data. By merging the excellent sensitivity of PET with the spatial resolution of CT, one can potentially improve tumor localization, especially in patients who have undergone therapy. Drawbacks include technical considerations, such as the quality of the CT scans, misregistration artifacts due to patient motion, and potentially imprecise measurements of the radioactivity concentration on CT attenuation-corrected PET scans. MRI Standard MRI

CT Angiography (CTA) CTA is now challenging catheter angiography as the primary method for assessing the vessels of the neck. In addition to evaluating the carotid arteries for evidence of atherosclerotic disease, CTA can effectively detect arterial dissection in trauma patients. Primary arterial disease, such as fibromuscular dysplasia, is also well demonstrated. Evaluation of vascular encasement by tumors is another application. Scanning protocols vary; however, thin-slice spiral data (e.g., 1-mm collimation and table speed of 2 mm/second) during a contrast bolus usually provide excellent images. Images derived from maximum intensity projection (MIP) and shaded surface display techniques can then be examined at a workstation. The advantages of CTA include its rapid data acquisition and relative noninvasiveness—properties that are especially important for the critically ill patient. Moreover, patients with cardiac pacemakers and ferromagnetic intracranial aneurysm clips, which are contraindications to MR angiography (MRA), can undergo CTA. The drawbacks of CTA include venous contamination due to ill-timed contrast boluses, physician time-intensive data manipulation at a workstation, artifacts due to metallic foreign bodies, and potentially confusing information caused by heavy mural calcification and adjacent bone. Additional Techniques Additional CT applications may prove useful. Perfusion CT (CTP), for example, is showing promise for evaluating masses by measuring the mean transit time, blood flow, and blood volume in benign neck lesions in comparison with malignant lesions (7). This application could potentially overcome some of the limitations of other techniques, such as positron emission tomography (PET). A drawback of CTP, however, is image deg-

MRI of the neck should be tailored for the anatomic region and process under evaluation. A standard head coil usually suffices for relatively localized examinations of the suprahyoid region and base of the skull. The infrahyoid neck requires a neck coil. Surface coils may improve the signal-to-noise ratio (SNR) by almost 50% compared to standard head coils; however, anatomic coverage may be limited. Axial, coronal, and sagittal sequences are essential. Unenhanced axial T1weighted images display anatomic relationships and can detect lesions (e.g., lymph node lesions) embedded within fat. T1-weighted coronal images can define the false vocal cords, true vocal cords, laryngeal ventricle, and the floor of mouth (10,11). T1-weighted sagittal images provide helpful information about the pre-epiglottic space and nasopharynx. T2-weighted transaxial images characterize tissue, detect tumor within muscle, demonstrate cysts, and assist differentiation of post-therapy fibrosis from recurrent tumor (12). Fast spin-echo (FSE) T2-weighted imaging has the added advantage of a relatively short acquisition time (13,14). Gradient moment nulling, flow compensation, cardiac gating, and presaturation pulses minimize motion artifacts. Phase and coding gradients are best oriented in the anterior-posterior direction to further reduce distracting artifacts (10). Gadolinium (Gd)-enhanced images improve delineation of margins in many lesions (15,16). Fat-suppression techniques, such as short tau inversion recovery (STIR) and frequency-selected fat suppression, may improve the conspicuity of soft-tissue lesions embedded in fatty tissue by selectively diminishing the hyperintensity of fat on T1-weighted images (15–17). Field inhomogeneity and artifact may partially diminish the potential advantages of this technique. Moreover, the normal enhancement of the aerodigestive mucosa may conceal small mucosal tumors.

CT and MRI in Head and Neck Imaging

455

Additional Techniques

Comparison of CT and MRI

Additional MR techniques may supplement standard protocols. For example, the magnetization transfer (MT) technique may be useful for differentiating enhancing lesions from background tissue and defining poorly enhancing lesions (18 –20). This technique relies on the transfer of magnetization between restricted protons associated with macromolecules and free protons of water to improve the tissue contrast. MT has not enjoyed widespread application in head and neck imaging, however, partly because conventional imaging usually provides sufficient delineation of most primary lesions and lymphadenopathy. Moreover, MT is associated with longer examination times and higher specific absorption rates (SARs) than conventional imaging. 3D reconstruction algorithms permit the 3D depiction of the tumor volume embedded within the surrounding soft tissues. 3D software and workstations with offline processing may facilitate the conceptual transformation of two-dimensional (2D) information into a 3D format and allow the examiner to “dissect” the display to a desired tissue depth (20).

MRI provides anatomic information on the neck that compares favorably with CT in most cases; however, its superiority to CT has not been conclusively established in all applications (11,25–32). The use of the multirow detector spiral technique has greatly enhanced the applications of CT, especially for vascular imaging, while advanced sequences and coil design have also improved MRI considerably over the past few years. Nevertheless, a distinct advantage of MRI is its superb soft-tissue contrast and multiplanar display by which blood vessels, masses, and adjacent soft tissues are easily differentiated (33,34). MR scanning is especially helpful for patients in whom the distinction between a mass and surrounding soft-tissue structures on CT is poor. MRI can usually define these lesions in three orthogonal planes, without the need to reposition the patient. Although CT scanning and its anatomic display have been primarily limited to the transverse plane, volume scanning using multirow detector spiral technique provides isotropic voxels and has significantly expanded the flexibility of multiplanar reformatted CT imaging. MRI better displays the lower neck without the degradation from shoulder artifact that occurs with CT. Dental restorations and densely calcified or ossified cartilages usually do not significantly degrade MR images. Although they are better tolerated than older agents, the newly developed intravenously administered, nonionic, iodinated contrast agents used in CT carry a definite risk of adverse reactions, including anaphylaxis, and may compromise marginal renal function (35). The noniodinated Gd compounds used in MRI are generally considered somewhat safer for individuals with impaired renal function or a history of reactions to contrast material; however, contrast reactions do occur even in MR scanning, and one should always exercise caution when administering contrast to any highly allergic patient (36). With MRI, motion artifacts from breathing, carotid artery pulsations, and swallowing may degrade images (10,37,38). Moreover, many patients with neck pathology have comorbidities, such as chronic obstructive pulmonary disease (COPD), or difficulty in controlling secretions, which may become issues during long examinations (10). Spiral CT scans can be obtained within minutes, which alleviates many of these problems in sick patients (39). Patients with cardiac pacemakers and metallic cochlear implants are restricted from the MR scanner because of the effect of the strong magnetic field and radiofrequency (RF) on these devices (40). Strongly ferromagnetic cerebral aneurysm clips and dorsal column stimulators may potentially dislodge or malfunction (41). Iron debris within the eye may cause visual impairment when subjected to the magnetic field (42). Dentures and dental amalgams may create artifacts, though generally they are not as imagedegrading as those seen in CT scans. Open-magnet design and newer scanner geometry coupled with appropriate levels of sedation have largely eliminated problems posed by anxiety and claustrophobia.

MR Spectroscopy (MRS) MRS has been used to examine 31-phosphorus, 19fluorine, and 13-carbon in tissues; however, most investigations have focused on hydrogen (proton) spectra. Although intracranial applications of MRS have been steadily increasing, applications in the extracranial head and neck have been disappointing. First of all, MRS requires a homogeneous magnetic field. Susceptibility artifacts introduced by the paranasal sinuses, airway, and bone, and pulsation artifacts from the carotid artery severely degrade data. Additionally, a large amount of fat within the neck produces a lipid peak that obscures the relatively small peaks of tumor markers, such as choline. Finally, MRS remains rather nonspecific. Because many potentially lethal head and neck tumors begin as small nests of cells within the mucosa, the lack of sufficient spatial resolution by MRS severely limits its application. Moreover, most head and neck lesions begin as mucosal processes, and the mucosa abuts the airway, further contributing to inhomogeneity and artifacts (21). Nevertheless, work continues and newer techniques may eventually make MRS practical for head and neck applications (22). Open-Bore and High-Field-Strength Magnets Although open-bore magnets may be used for the obese or claustrophobic patient, anatomic depiction of complex anatomy and disease processes may be disappointing. These units have been especially useful for biopsies and innovative procedures such as photodynamic laser therapy (23,24). Finally, the advent of clinically available 3T magnets creates both opportunities and challenges for head and neck imaging. The exaggerated chemical shift poses the potential of heightened artifacts in the fat-laced anatomy of the neck; however, the improved signal and contrast may be exploited for better images.

456

Wippold

Table 1 Indications for Imaging by Anatomic Region Anatomic region Orbit Skull base

Paranasal sinuses Deep spaces Larynx Lymph nodes

CT

MR

Trauma; acute infection Trauma; conductive hearing loss; acquired middle ear cholesteatoma; bone erosion Trauma; infection

Masses; complicated infection, i.e., intracranial involvement Masses (especially suspected intracranial or neck extension); neurosensory hearing loss; perineural spread of tumor Cephalocele; complicated infection; masses (especially suspected intracranial or orbit extension) Masses; perineural spread of tumor Masses (CT and MR imaging complementary); posttherapy evaluation Lymphadenopathy (CT and MR imaging complementary)

Acute infection; sialolithiasis Trauma; masses (CT and MR imaging complementary) Lymphadenopathy (CT and MR imaging complementary)

APPROPRIATE APPLICATIONS Orbit MRI has become the modality of choice for evaluating nontraumatic orbit pathology (Table 1). Superior softtissue contrast, multiplanar display, and fat-saturation enhancement techniques combine to produce a powerful analytic tool, especially for defining mass lesions. Acute traumatic injury and some infections remain within the realm of CT imaging, primarily because of the availability of scanners and the speed of scanning, which is advantageous for seriously ill patients. Small calcifications and densities produced by foreign bodies and tumors, such as meningiomas, are also best detected with CT. For the orbit, axial scans without and with the use of intravenous contrast material provide an adequate baseline evaluation. Bone algorithm reconstructions are extremely useful in cases of trauma. Coronal images are also very useful, especially for evaluating extraocular muscle caliber, optic nerves, paranasal sinuses, and the cribriform plate. An effective way to organize the anatomy and pathology of the orbit is to divide this region into distinct compartments based on fascial layers that are easily defined with cross-sectional imaging. These regions often have distinct pathologic differential diagnoses. The compartment containing the globe itself is encased by Tenon’s capsule. Tenon’s capsule in turn separates the globe from extraocular tissues. The myofascial sling invests the extraocular muscle cone and separates the intraconal space, which contains fat and the optic nerve, from the extraconal space. The periorbital fascia separates the intraorbital extraconal space from extraorbital anatomy. The globe consists of the fluid-filled anterior chamber, posterior chamber, and vitreous body. Because of the superb tissue differentiation capabilities of MR scanning, hemorrhage and masses can be easily detected and characterized when they are sufficiently large. Smaller lesions are best clinically evaluated with fundoscopy; however, MR scanning may provide valuable information about scleral penetration. Fluid collections, such as blood within Tenon’s space, are also easily diagnosed with MRI. Fat-saturated fluid-attenuated inversion recovery (FLAIR) imaging and diffusionweighted imaging (DWI) may be useful for following therapeutic response to inflammation (43). Choroidal, hyaloid, and retinal detachments may be differentiated with MRI.

The intraconal space may be further divided into the optic nerve and intraconal fat. The optic nerve is best studied with MR scanning in the coronal plane supplemented with axial and sagittal plane information. FLAIR imaging may detect hyperintensity in various conditions, such as optic neuritis, and the addition of T1-weighted Gd-enhanced sequences, especially when coupled with a fat-saturation technique, may define active inflammation or tumors (Fig. 1). Increases in the apparent diffusion coefficient (ADC) in DWI have been observed in chronic, postinflammatory optic nerve lesions in demyelination (44). The fat within the intraconal space provides excellent inherent signal contrast properties for detecting masses, such as hemangiomas and schwannomas. Care must be taken when interpreting fat-saturation sequences because of the potentially misleading effect of field inhomogeneity. The imager should always review non-fat-suppressed T1-weighted and T2-weighted images to verify inflammatory infiltrates within this space. Advances in phase-corrected algorithms for three-point Dixon (3PD) imaging, in which pure water, pure fat, and water-plus-fat online image reconstructions are produced, promise to increase the SNR and contrast-to-noise ratio (CNR) (45). The extraocular muscle cone is nicely silhouetted by intraconal fat. MRI outlines enlarged muscles and the tendonous insertions. Coronal images are excellent for verifying muscle size. Reconstructions derived from magnetization-prepared rapid-gradient-echo (MPRAGE) images can be used to correct difficulties in patient positioning and can facilitate complex multiplanar demonstrations of difficult pathology. Because of the lack of signal intensity from cortical bone, the optic nerve within the optic canal can be well visualized. With the use of Gd enhancement, subtle meningeal pathology can be detected. Paranasal sinus pathology, which can extend into the extraconal space, is also well characterized with MRI. The use of signal intensity on T2weighted images and enhancement properties can assist in differentiating tumor spread from inflammatory disease. Skull Base Most studies divide the skull base into anterior, central, and posterior regions. Conceptually, the anterior skull base is dominated by the ethmoid bone, the central

CT and MRI in Head and Neck Imaging

457

skull base is dominated by the sphenoid bone, and the posterior skull base is dominated by the occipital bone. Similarly, lesions can be grouped into those that originate above the skull base and invade inferiorly, those that involve the bone itself, and those that begin inferior to the skull base and invade superiorly (46,47). MRI clearly excels in evaluating the skull base, especially lesions with intracranial and neck soft-tissue components (Table 1, Fig. 2). Although CT does provide valuable information about cortical bone itself, the sensitivity of contrast-enhanced MR scans combined with multiplanar capabilities for detecting meningeal spread and cranial nerve involvement clearly vaults MRI into the dominant role. CT typically requires a bony algorithm for fine-bone detection, which further diminishes its value for evaluating the adjacent brain. For skullbase evaluations with CT, axial scans are typically used. The spiral technique provides zoomed 1-mm high-resolution axial slices that can then be reformatted in multiple planes with excellent spatial resolution.

Figure 2. Meningioma arising in the anterior skull base. Sagittal enhanced T1-weighted image demonstrates a large, markedly enhancing, extra-axial soft-tissue mass within the frontal lobes. The tumor has extended below the cribriform plate into the ethmoid paranasal sinus, and has also invaded the sella turcica perched above the pituitary gland.

Figure 1. Retrobulbar neuritis in a patient with multiple sclerosis. a: Coronal enhanced T1-weighted MR image with fat saturation demonstrating an enlarged enhancing right optic nerve. b: Corresponding axial FLAIR T2-weighted MR image showing the hyperintense periventricular demyelinating plaques.

Soft-tissue algorithms usually require somewhat thicker slices (e.g., 5 mm), and may be supplemented with intravenous contrast material. MRI, on the other hand, can fully evaluate the brain. The flow void of the carotid arteries also provides excellent inherent contrast with tumors that potentially encase such vessels. High-resolution sequences, such as 3D constructive interference in the steady state (CISS), give the appearance of a heavily T2-weighted sequence and can be used to examine skull-base anatomy in regions such as the cavernous sinus (48). The temporal bone can also be included in skull-base studies. For conductive hearing loss, CT scanning remains the initial screening modality because of its excellent depiction of the middle-ear ossicles and bony anatomy (Fig. 3). The spiral technique with submillimeter reconstruction yields detail never before experienced by imagers. Inflammatory temporal bone pathology is also best studied with CT. For neurosensory hearing loss, however, MRI offers a variety of techniques to fully evaluate the internal auditory canal and cerebellopontine angle. Traditional high-resolution, thin-slice, contrast scanning can detect enhancing schwannomas within the internal auditory canal (Fig. 4). High-resolution T2-weighted images using 3D fast recovery FSE (FRFSE) and CISS imaging can dissect individual nerves without the use of contrast, and yield impressive 3D renderings (49,50) (Fig. 5). Diagnoses such as cochlear nerve atrophy, cochlear nerve aplasia, and endolymphatic sac lesions are now within the reach of imaging. DWI has also been proposed for differentiation of recurrent cholesteatoma from granulation tissue after mastoidectomy (51). MRI has also made it possible to confidently diagnosis of perineural spread of malignancies (52–55). A diagnosis of perineural spread of skull-base lesions is critically important for treatment planning, and indicates far more extensive disease than may be clinically detected by even the most astute surgeons. CT usually only indirectly infers perineural spread by demonstrating enlargement of the bony foramina. Fat-saturation

458

Wippold

Figure 5. Normal cochlea. Volume-rendered 3D oblique views of the cochlea are based on 3D T2-weighted images acquired at 3T. (Case courtesy of Benjamin C.P. Lee.)

Figure 3. Superior semicircular canal dehiscence. Reformatted axial CT images demonstrate lack of bony covering of the superior semicircular canal (arrows) consistent with dehiscence. CT scanning is the preferred method for evaluating conductive hearing loss and the fine bony anatomy of the temporal bone.

Gd-enhanced MR scans are often capable of detecting subtle tumor tracking along the Vth and VIIth cranial nerves, as well as other nerves that travel through the many foramina of the skull base, before the lesions have grown sufficiently large to affect the surrounding bone (Fig. 6). Paranasal Sinuses CT remains the acknowledged modality of choice for the routine evaluation of paranasal sinus inflammatory disease because of its excellent depiction of bone (56) (Table 1, Fig. 7). Direct coronal scanning using a bone algorithm and 1- to 3-mm-thick slices are recom-

Figure 4. Vestibular schwannoma. Coronal enhanced T1weighted MR image shows an enhancing right intracanalicular mass (arrowhead). MRI is preferred for evaluation of neurosensory hearing loss.

mended for the paranasal sinuses because of the superior portrayal of the osteomeatal units in this plane. MRI can complement CT for diagnostically challenging cases. For example, MRI may reveal a cephalocele filled with characteristic brain and cerebrospinal fluid signals eroding the cribriform plate, whereas CT may not be able to differentiate cephalocele from other causes of soft-tissue density, such as nasal polyp or inflammatory disease (57). Additionally, MRI provides important signal information that may differentiate tumor from inflammatory disease. As a general rule, most inflammatory diseases have hyperintense signal characteristics on T2-weighted images. As secretions inspissate, signal may change on both T1- and T2-weighted images. Careful clinical correlation is therefore essential in such cases (58). Because mucosal secretions are avascular, they will typically fail to enhance after the administration of Gd. The inflamed mucosa will enhance in a ring configuration that lines the sinus cavity. Tumors, on the other hand, usually display intermediate signal intensity on T2-weighted images and homogeneously enhance (Fig. 8). Exceptions to these rules

Figure 6. Extensive skull-base squamous cell carcinoma with perineural spread of tumor involving the trigeminal and facial nerves. Axial enhanced T1-weighted image demonstrates enhancement of the left facial nerve. Note the atrophy of the left muscles of mastication.

CT and MRI in Head and Neck Imaging

459

Figure 7. Sinonasal polyposis. Coronal CT scan shows a soft-tissue mass filling the upper portion of the nasal cavity and ethmoid sinuses with bilateral extension through the osteomeatal units into the medial portions of the maxillary sinuses. Minimal mucosal thickening is identified in the inferior portions of both maxillary sinuses. The left frontal sinus is also opacified.

include tumors in which necrotic foci fail to enhance, and adenoid cystic carcinomas in which the often markedly hyperintense signal on T2-weighted images may mimic inflammatory disease. MRI is also ideally suited for evaluating anterior cranial fossa involvement of paranasal sinus tumors that have perforated the cribriform plate (Figs. 2 and 8). Subtle tell-tale meningeal enhancement may be easily detected by coronal sequences. Orbital involvement by paranasal sinus tumors is also well evaluated with coronal and axial MRI using Gd intravenous contrast and fat-saturation techniques.

Figure 8. Squamous cell carcinoma of the ethmoid paranasal sinus. Coronal enhanced T1-weighted MR image with fat saturation demonstrates a homeogeneously enhancing left ethmoid soft-tissue mass. The mass extends inferiorly into the superior nasal cavity and invades the left maxillary sinus through the osteomeatal unit. The mass has perforated the cribriform plate and invaded the anterior cranial fossa. The left orbit has also been invaded. The tumor has extended across the midline to involve the right ethmoid sinus.

Deep Spaces of the Neck With the advent of cross-sectional imaging, the traditional cervical triangle method of organizing the anatomy of the neck has been largely replaced by a system that defines neck spaces according to the cervical fascial planes. The differential diagnoses of diseases of the neck are related to the space of origin, and therefore this system of relying on the fascial planes is ideally suited to both imaging and clinical management of head and neck patients. The cervical spaces of the suprahyoid and infrahyoid neck include the sublingual space, submandibular space, buccal space, parotid space, parapharyngeal space, carotid space, masticator space, pharyngeal mucosal space, visceral space, retropharyngeal space, posterior cervical space, and perivertebral space. The paired sublingual spaces are located in the floor of the mouth and are defined by the mandible anteriorly and laterally, the hyoid bone posteriorly, the oral mucosa superiorly, and the mylohyoid muscle inferiorly (59). Contained within this space are the paired sublingual salivary glands; the deep lobes of the submandibular salivary glands; submandibular ducts; the genioglossus, geniohyoid, hyoglossus, and styloglossus muscles; the lingual arteries; the lingual branches of the third division of the Vth cranial nerve; the glossopharyngeal and hypoglossal nerves; and fat (60). Typical lesions include carcinomas extending from the floor of the mouth and the tongue, ranulas, dermoids and epidermoids, hemangiomas and lymphangiomas, lingual thyroid glands and thyroglossal duct cysts, abscesses, enlarged lymph nodes, calculi within the submandibular duct, and schwannomas (60,61). MRI, with its multiplanar capabilities and fat-saturation techniques, is ideally suited for studying this space because of the lack of significant artifacts from dental restorations and beam-hardening. CT remains the study of choice for evaluating calculi (Table 1). The submandibular space is posterolateral to the sublingual space and contains the superficial lobe of the submandibular salivary gland and lymph nodes (59). Lesions within this space include cystic hygromas, branchial cleft cysts, dermoids, epidermoids, thyroglossal duct cysts, abscesses, calculi from the submandib-

460

Figure 9. Hemangioma of the left parotid gland. A large hyperintense intraparotid hemangioma is identified on axial T2weighted MR image. The lesion invades the masticator space anteriorly and the parapharyngeal space medially.

ular glands, and salivary gland masses. Because of the high fat content of this space and the limited variety of structures found within it, both MRI and CT are routinely used for evaluations. MRI is advantageous for characterizing tissue in neoplasms, and CT is preferred for patients with acute infections. The buccal space is a small region anterior to the masseter and lateral to the buccinator muscle. Pathologic conditions that occur in this space include infection, deeply invasive skin cancers, minor salivary gland tumors, and facial lymphadenopathy (62– 64). MRI best evaluates this space. The parotid space is located posterior to the masseter muscle and contains the parotid gland, Stenson’s duct, the facial nerve, and intraparotid lymph nodes and blood vessels (59,65). Salivary lesions, such as pleomorphic adenoma, mucoepidermoid carcinoma, and adenoid cystic carcinoma, make up the vast majority of lesions within this space. Lymphadenopathy and hemangiomata may also be seen. Because of the subtle water and fat contrast within the parotid gland and the surrounding tissues, MRI has become the method of choice for evaluating masses (66,67) (Fig. 9). MRI is also excellent for patients with suspected autoimmune disease, such as Sjo¨gren’s syndrome, and some studies have examined the use of MR sialography in such conditions (68 –70). CT is reserved for stone disease and some acute infections. The parapharyngeal space extends from the skull base to the hyoid bone and is posteromedial to the masticator space. The parapharyngeal space is subdivided into prestyloid and poststyloid compartments. The prestyloid compartment contains branches of the internal maxillary and ascending pharyngeal arteries, nerves, fat, salivary rests, and minor salivary glands (62). Primary lesions of the prestyloid compartment are often salivary in origin, such as pleomorphic adenomas, but other lesions, such as carcinomas invading from adjacent spaces, lipomas, and schwannomas, also occur with relative frequency (71). The poststyloid compartment (also known as the carotid space) contains the carotid artery, internal jugular vein, glossopharyn-

Wippold

geal nerve, vagus nerve, spinal accessory nerve, hypoglossal nerve, sympathetic chain, and internal jugular lymph nodes (71,72). Lesions occurring in this space arise from the tissues found within this region and notably include glomus tumors and schwannomas (73). MRI is the preferred method for evaluating the parapharyngeal space. The masticator space is formed by the splitting of the superficial layer of the deep cervical fascia, and includes the mandible, the muscles of mastication, and the mandibular division of the trigeminal nerve (59,65,71,74). Lesions derived from these tissues include nerve sheath tumors, mandibular and soft-tissue sarcomas, dental tumors, cysts and abscesses, osteomyelitis, and lipomas (75,76). Vascular lesions, such as hemangiomas, lymphangiomas, and, rarely, pseudoaneurysms, may also occur. In children, rhabdomyosarcoma also may present in the masticator space. The foramen ovale at the skull base forms the “chimney of the masticator space” and is a frequent conduit of lesions traversing the masticator space and middle cranial fossa. Although CT demonstrates the cortex of the mandible in acutely traumatized or infected patients, MRI is the modality of choice for neoplasms. Multiplanar acquisitions and Gd contrast-enhanced images best demonstrate perineural spread involving the trigeminal nerve. The pharyngeal mucosal space includes the mucosal surfaces and immediate submucosa of the nasopharynx and oropharynx. The oral cavity and suprahyoid portion of the hypopharynx can also be included in this space (77). Mucosa, lymphoid tissue, minor salivary glands, pharyngeal constrictor muscles, and the salpingopharyngeus muscle are contained in this space (77,78). Predictably, the most common malignant lesion of the pharyngeal mucosal space is carcinoma, although benign lesions, such as salivary tumors, may also occur. Lesions may be superficial, in which case MR scanning may yield false-negative results, or deeply invasive, in which case MRI can be very helpful. Although MRI is generally considered the best choice for evaluating mandibular involvement with oral cavity carcinoma, chemical shift artifact and concomitant inflammation may diminish its specificity (79,80). Methods of tumor volume assessment have also received recent attention, especially with oncology therapy protocols requiring objective measurements of therapy response. Both CT and MR scans have been used in such determinations, although validation continues to be an issue (81). The visceral space contains the larynx and hypopharynx, the thyroid and parathyroid glands, the trachea and esophagus, peritracheal lymph nodes, and the recurrent laryngeal nerves (65). Lesions typically reflect these tissues and include squamous cell carcinomas, thyroid and parathyroid adenomas and carcinomas, and schwannomas. The retropharyngeal space lies posterior to the visceral space and extends from the base of the skull to the mediastinum. This space serves as a conduit for the spread of neck pathology into the chest (82,83). Lymph nodes occupy the suprahyoid portion of this space and may enlarge in oropharyngeal infection and malignan-

CT and MRI in Head and Neck Imaging

cies. CT tends to serve as the first line of diagnosis, especially in infection; however, it is difficult to differentiate between retropharyngeal abscess and adenitis because both processes cause hypodense regions within the inflammatory mass. MRI is excellent for detecting metastatic retropharyngeal lymph nodes (84) (Fig. 10). The posterior cervical space abuts the carotid space posterolaterally and is sandwiched by the sternocleidomastoid muscle anterolaterally and the paraspinal muscles posteromedially (59,85). The primary components of this space are fat, the spinal accessory and dorsal scapular nerves, and the spinal accessory lymph nodes of the deep cervical chain. Among the lesions most commonly seen in this space is lymphadenopathy from metastatic squamous cell carcinoma, with primary sites in the pharyngeal mucosal space and lymphoma. Lipomas, sarcomas, cystic hygromas, and branchial cleft cysts are also seen in this space. Because fat, the major component of this space, provides naturally occurring tissue contrast, both CT and MRI perform well. Large extracapsular nodal masses tend to blend with the muscle boundaries of this space, however, and MRI is superior to CT in this application. The perivertebral space is formed by the deep cervical fascia. An anterior compartment envelops the vertebral bodies and spinal cord, the vertebral arteries, phrenic nerve, and perivertebral and scalene muscles. A posterior compartment contains the posterior vertebral elements and the paraspinous muscles (59,65,83). Examples of pathology include vertebral osteomyelitis, necrotizing fasciitis and metastases, nerve sheath tumors (e.g., plexiform neurofibromas), and rare lesions (e.g., chordoma). MRI is preferred for depicting soft-

Figure 10. Adenoid cystic carcinoma. Axial T1-weighted enhanced MR image demonstrates a large hyperintense adenoid cystic carcinoma of the nasal cavity invading the left maxillary sinus, left masticator space, left parapharyngeal space, posterior right maxillary sinus, right masticator space, and right parapharyngeal space. Bilateral retropharyngeal node metastases are evident.

461

Figure 11. Supraglottic squamous cell carcinoma of the larynx. Axial contrast-enhanced CT scan reveals a thickened epiglottis that is more prominent on the right. The surrounding paraglottic fat is not involved. A necrotic right jugular chain lymph node is also noted.

tissue lesions, although CT remains very useful for precisely identifying details of bone destruction. Larynx The larynx is usually anatomically divided into the supraglottic, glottic, and subglottic (infraglottic) regions. The supraglottic division is defined as the portion of the larynx that extends from the superior-most tip of the epiglottis to a transverse plane through the laryngeal ventricle. The glottis extends from this transverse plane to 1 cm inferiorly and includes the true vocal cords. The subglottic region extends from the inferior-most plane of the true cords to the inferior portion of the cricoid cartilage. The vast majority of lesions discovered in patients referred for imaging of the larynx are squamous cell carcinomas, although benign lesions, such as laryngoceles, do occasionally appear (86,87). Small benign entities, such as polyps, are usually easily diagnosed in the clinician’s office and typically are not scanned. Multiplanar MRI is especially well suited for analysis of laryngeal lesions (88). Soft-tissue contrast aids in the early detection of spread of tumor into paraglottic and preepiglottic fat. These areas are not easily evaluated by clinical endoscopy, and have a negative predictive value for survival. The use of Gd intravenous contrast material further refines detection (89). Cartilage invasion, carotid encasement, and transglottic involvement via submucosal spread are also well studied with MR scanning (27,88,90). Because spiral CT is readily available and scanning times are extremely rapid, CT remains a valuable and frequently used screening modality for the larynx (Table 1, Fig. 11). CT is also attractive for patients undergoing imaging of the larynx, because many of these patients have COPD and/or difficulty with secretions. Although the value of MRI for imaging the post-treatment larynx exceeds that of CT, false-negative or false-

462

Wippold

positive interpretations may still result. PET, especially when coupled with CT, is a useful adjunct. Lymph Nodes Cervical lymphatics are critical conduits that drain the skin and mucosal surfaces of the head and neck. Nearly 40% of all lymph nodes in the body are located above the clavicles; only the orbit and cervical muscles do not have direct lymphatic connections. Lymph nodes are usually embedded within the fat planes that surround the vessels and separate major cervical muscles. Therefore, the fat of the neck provides an excellent natural contrast with the nodes on T1-weighted MR images. Lymph nodes are divided into 10 major groups (59,91). These groups are named for the structures in proximity to nodal location. Six groups (the occipital, mastoid, parotid, submandibular, facial, and submental groups) form a sentinel ring at the skull base. Additionally, a sublingual group lies deep in the floor of the mouth, and the retropharyngeal group is embedded within the retropharyngeal space. These sentinel groups drain into paired anterior and lateral cervical chains to extend from the skull base to the clavicles. Of these groups, the submental, submandibular, retropharyngeal, and lateral cervical chains play especially important roles in the spread of head and neck disease. The lateral cervical chain is subdivided into the superficial and deep lateral cervical lymph nodes. The superficial group follows the course of the external jugular vein. It is usually palpable, and therefore is not usually examined by imaging. The important deep group is further divided into the spinal accessory, transverse cervical, and internal jugular groups. The spinal accessory nodes are found within the fat of the posterior cervical space adjacent to the spinal accessory nerve, and the internal jugular group tracks along the internal jugular vein. The American Joint Committee on Cancer (AJCC) and the American Academy of Otolaryngology—Head and Neck Surgery divide these 10 lymph node chains into a series of levels that have prognostic importance (92,93). Briefly, level 1 consists of the submental and submandibular nodes. Level 2 includes the internal jugular nodes extending from the base of the skull to the carotid bifurcation (hyoid bone). Normal nodes in levels 1 and 2 may measure up to 1.5 cm in diameter. Level 3 corresponds to the internal jugular nodes from the carotid bifurcation to the omohyoid muscle (cricoid cartilage). Level 4 refers to all nodes in the internal jugular group from the omohyoid muscle to the clavicle. Level 5 consists of the spinal accessory and transverse cervical nodes that occupy the posterior cervical triangle. Level 6 contains the pretracheal, prelaryngeal, and paratracheal nodes. Finally, level 7 includes the nodes in the tracheoesophageal groove and upper mediastinum. Nodes from level 3 to level 7 should not exceed a maximum diameter of 1 cm. Nodal enlargement beyond the acknowledged maximum diameter is one of the most common signs of abnormality, although false-negative and false-positive results may occur. Necrosis also indicates abnormality. Clusters of three or more contiguous ill-defined nodes

Figure 12. Metastatic lymphadenopathy. Axial contrast-enhanced CT scan shows large bilateral necrotic jugular chain lymph nodes spread from a laryngeal primary site.

within the same level ranging from 8 mm to 15 mm in size may also be considered abnormal. Although nodal shape is no longer thought to be a reliable sign for differentiating normal from pathologic nodes, round nodes tend to be neoplastic, whereas elliptical or beanshaped nodes are generally normal or hyperplastic (92,94). Despite these criteria, a large number of falsenegative and false-positive interpretations may result. Moreover, recognition of nodal abnormality on crosssectional imaging does not necessarily guarantee identification of etiology. Extensive adjacent soft-tissue stranding coupled with a rapidly developing history of fever and painful adenopathy, implies inflammation, whereas a subacute or chronic clinical course, a slowly growing relatively painless mass that is hard to palpation, and nodal necrosis on imaging suggest neoplasm. Lymphadenopathy arising from squamous cell carcinoma often displays necrosis, whereas nodes associated with lymphoma typically do not. The use of ADC values and dynamic contrast-enhanced (DCE) MR scans may further assist in determining etiology (95– 97). The imaging community is divided in its recommendations regarding an appropriate modality for evaluating lymphadenopathy (26,32,98). CT remains very popular because of it availability, speed, and excellent spatial resolution. Lymph nodes are usually embedded within fat, and fat is well portrayed by CT (Figs. 11 and 12). Alternatively, MRI has superior soft-tissue contrast and multiplanar capabilities (Table 1, Fig. 10). In an attempt to further refine the diagnostic power of MRI, dextran-coated ultrasmall superparamagnetic iron oxide (USPIO) has been developed as a lymph node contrast agent based on the siderocytic capabilities of normal nodes compared with pathologic nodes. However, technical factors, such as 24- to 36-hour imaging times, have limited its widespread application (99,100). CONCLUSIONS Cross-sectional imaging has revolutionized the practice of head and neck radiology, as well as the clinical specialty of otolaryngology/head and neck surgery. The

CT and MRI in Head and Neck Imaging

complementary roles of CT and MRI offer powerful tools for the diagnosis and management of benign and malignant conditions, provided that the radiologist understands the clinical issues regarding the patient and the anatomy under investigation. With the advent of 3T magnets and innovative pulse sequences and coils, the field of head and neck imaging will continue to evolve in the years to come.

REFERENCES 1. Gatenby RA, Mulhern Jr CB, Strawitz J, Moldofsky PJ. Comparison of clinical and computed tomographic staging of head and neck tumors. AJNR Am J Neuroradiol 1985;6:399 – 401. 2. Stark DD, Moss AA, Gamsu G, Clark OH, Gooding GA, Webb WR. Magnetic resonance imaging of the neck. Part II: Pathologic findings. Radiology 1984;150:455– 461. 3. Keberle M, Kenn W, Hahn D. Current concepts in imaging of laryngeal and hypopharyngeal cancer. Eur Radiol 2002;12:1672– 1683. 4. Spreer J, Krahe T, Jung G, Lackner K. Spiral versus conventional CT in routine examinations of the neck. J Comput Assist Tomogr 1995;19:905–910. 5. Mukherji SK, Castillo M, Huda W, et al. Comparison of dynamic and spiral CT for imaging the glottic larynx. J Comput Assist Tomogr 1995;19:899 –904. 6. Suojanen JN, Mukherji SK, Wippold FJ. Spiral CT of the larynx. AJNR Am J Neuroradiol 1994;15:1579 –1582. 7. Rumboldt Z, Al-Okaili R, Deveikis JP. Perfusion CT for head and neck tumors: pilot study. AJNR Am J Neuroradiol 2005;26:1178 – 1185. 8. Emami B, Purdy JA, Simpson JR, Harms W, Gerber R, Wippold JF. 3-D conformal radiotherapy in head and neck cancer. The Washington University experience. Front Radiat Ther Oncol 1996;29: 207–220. 9. Yousem DM, Sack MJ, Scanlan KA. Biopsy of parapharyngeal space lesions. Radiology 1994;193:619 – 622. 10. Hudgins PA, Gussack GS. MR imaging in the management of extracranial malignant tumors of the head and neck. AJR Am J Roentgenol 1992;159:161–169. 11. Jabour BA, Lufkin RB, Hanafee WN. Magnetic resonance imaging of the larynx. Top Magn Reson Imaging 1990;2:60 – 68. 12. Glazer HS, Lee JK, Levitt RG, et al. Radiation fibrosis: differentiation from recurrent tumor by MR imaging. Radiology 1985;156: 721–726. 13. Lewin JS, Curtin HD, Ross JS, Weissman JL, Obuchowski NA, Tkach JA. Fast spin-echo imaging of the neck: comparison with conventional spin-echo, utility of fat suppression, and evaluation of tissue contrast characteristics. AJNR Am J Neuroradiol 1994; 15:1351–1357. 14. Yousem DM, Hurst RW. MR of cervical lymph nodes: comparison of fast spin-echo and conventional spin-echo T2W scans. Clin Radiol 1994;49:670 – 675. 15. Hasso AN, Brown KD. Use of gadolinium chelates in MR imaging of lesions of the extracranial head and neck. J Magn Reson Imaging 1993;3:247–263. 16. Vogl T, Dresel S, Juergens M, Assal J, Lissner J. MR imaging with Gd-DTPA in lesions of the head and neck. J Otolaryngol 1993;22: 220 –230. 17. Barakos JA, Dillon WP, Chew WM. Orbit, skull base, and pharynx: contrast-enhanced fat suppression MR imaging. Radiology 1991; 179:191–198. 18. Balaban RS, Ceckler TL. Magnetization transfer contrast in magnetic resonance imaging. Magn Reson Q 1992;8:116 –137. 19. Gillams AR, Fuleihan N, Grillone G, Carter AP. Magnetization transfer contrast MR in lesions of the head and neck. AJNR Am J Neuroradiol 1996;17:355–360. 20. Lofchy NM, Stevens JK, Brown DH. Three-dimensional imaging of the parapharyngeal space. Arch Otolaryngol Head Neck Surg 1994;120:333–336. 21. Mukherji SK, Schiro S, Castillo M, Kwock L, Muller KE, Blackstock W. Proton MR spectroscopy of squamous cell carcinoma of the extracranial head and neck: in vitro and in vivo studies. AJNR Am J Neuroradiol 1997;18:1057–1072.

463 22. King AD, Yeung DKW, Ahuja AT, Leung SF, Tse GMK, van Hasselt AC. In vivo proton MR spectroscopy of primary and nodal nasopharyngeal carcinoma. AJNR Am J Neuroradiol 2004;25:484 – 490. 23. Jager HR, Taylor MN, Theodossy T, Hopper C. MR imagingguided interstitial photodynamic laser therapy for advanced head and neck tumors. AJNR Am J Neuroradiol 2005;26:1193– 1200. 24. Lambre H, Anzai Y, Farahani K, Castro D, Lufkin RB. Interventional magnetic resonance imaging of the head and neck and new imaging techniques. Neuroimaging Clin N Am 1996;6:461– 472. 25. Curtin HD. Imaging of the larynx: current concepts. Radiology 1989;173:1–11. 26. Madison MT, Remley KB, Latchaw RE, Mitchell SL. Radiologic diagnosis and staging of head and neck squamous cell carcinoma. Radiol Clin North Am 1994;32:163–181. 27. van den Brekel MW, Castelijns JA, Snow GB. The role of modern imaging studies in staging and therapy of head and neck neoplasms. Semin Oncol 1994;21:340 –348. 28. Phelps PD. Carcinoma of the larynx—the role of imaging in staging and pre-treatment assessments. Clin Radiol 1992;46: 77– 83. 29. van Dijke CF, van Waes PF. Head and neck tumors, MRI versus CT: a technology assessment pilot study. Eur J Radiol 1992;14: 235–239. 30. Som PM. Detection of metastasis in cervical lymph nodes: CT and MR criteria and differential diagnosis. AJR Am J Roentgenol 1992; 158:961–969. 31. Swartz JD, Yussen PS, Popky GL. Imaging the soft tissues of the neck: non-nodal acquired disease. Crit Rev Diagn Imaging 1991; 31:471–547. 32. Yousem DM, Som PM, Hackney DB, Schwaibold F, Hendrix RA. Central nodal necrosis and extracapsular neoplastic spread in cervical lymph nodes: MR imaging versus CT. Radiology 1992; 182:753–759. 33. Hoover LA, Wortham DG, Lufkin RB, Hanafee WN. Magnetic resonance imaging of the larynx and tongue base: clinical applications. Otolaryngol Head Neck Surg 1987;97:245–256. 34. Sakai F, Gamsu G, Dillon WP, Lynch DA, Gilbert TJ. MR imaging of the larynx at 1.5 T. J Comput Assist Tomogr 1990;14: 60 –71. 35. Harris KG, Smith TP, Cragg AH, Lemke JH. Nephrotoxicity from contrast material in renal insufficiency: ionic versus nonionic agents. Radiology 1991;179:849 – 852. 36. Haustein J, Niendorf HP, Krestin G, et al. Renal tolerance of gadolinium-DTPA/dimeglumine in patients with chronic renal failure. Invest Radiol 1992;27:153–156. 37. Wippold FJ. Diagnostic imaging of the larynx. In: Cummings CW, Flint PW, Haughey BH, et al., editors. Otolaryngology— head and neck surgery. 4th ed. ed. St. Louis: C.V. Mosby; 2005. p 2026 – 2053. 38. Wippold FJ. Postoperative pharynx. In: Gore RM, Levine MS, editors. Textbook of gastrointestinal radiology. 2nd ed. Philadelphia: W.B. Saunders; 2000. p. 257–270. 39. Kalender WA, Vock P, Polacin A, Soucek M. [Spiral-CT: a new technique for volumetric scans. I. Basic principles and methodology]. Rontgenpraxis 1990;43:323–330. 40. Pavlicek W, Geisinger M, Castle L, et al. The effects of nuclear magnetic resonance on patients with cardiac pacemakers. Radiology 1983;147:149 –153. 41. Klucznik RP, Carrier DA, Pyka R, Haid RW. Placement of a ferromagnetic intracerebral aneurysm clip in a magnetic field with a fatal outcome. Radiology 1993;187:855– 856. 42. Kelly WM, Paglen PG, Pearson JA, San Diego AG, Soloman MA. Ferromagnetism of intraocular foreign body causes unilateral blindness after MR study. AJNR Am J Neuroradiol 1986;7:243– 245. 43. Rumboldt Z, Moses C, Wieczerzynski U, Saini R. Diffusionweighted imaging, apparent diffusion coefficients, and fluid-attenuated inversion recovery MR imaging in endophthalmitis. AJNR Am J Neuroradiol 2005;26:1869 –1872. 44. Hickman SJ, Wheeler-Kingshott CA, Jones SJ, et al. Optic nerve diffusion measurement from diffusion-weighted imaging in optic neuritis. AJNR Am J Neuroradiol 2005;26:951–956.

464 45. Rybicki FJ, Chung T, Reid J, Jaramillo D, Mulkern RV, Ma J. Fast three-point dixon MR imaging using low-resolution images for phase correction: a comparison with chemical shift selective fat suppression for pediatric musculoskeletal imaging. AJR Am J Roentgenol 2001;177:1019 –1023. 46. Laine FJ, Nadel L, Braun IF. CT and MR imaging of the central skull base. Part 1: Techniques, embryologic development, and anatomy. Radiographics 1990;10:591– 602. 47. Laine FJ, Nadel L, Braun IF. CT and MR imaging of the central skull base. Part 2. Pathologic spectrum. Radiographics 1990;10: 797– 821. 48. Yagi A, Sato N, Taketomi A, et al. Normal cranial nerves in the cavernous sinuses: contrast-enhanced three-dimensional constructive interference in the steady state MR imaging. AJNR Am J Neuroradiol 2005;26:946 –950. 49. Naganawa S, Koshikawa T, Fukatsu H, Ishigaki T, Aoki I, Ninomiya A. Fast recovery 3D fast spin-echo MR imaging of the inner ear at 3 T. AJNR Am J Neuroradiol 2002;23:299 –302. 50. Lane JI, Ward H, Witte RJ, Bernstein MA, Driscoll CLW. 3-T imaging of the cochlear nerve and labyrinth in cochlear-implant candidates: 3D fast recovery fast spin-echo versus 3D constructive interference in the steady state techniques. AJNR Am J Neuroradiol 2004;25:618 – 622. 51. Maheshwari S, Mukherji SK. Diffusion-weighted imaging for differentiating recurrent cholesteatoma from granulation tissue after mastoidectomy: case report. Am J Neuroradiol 2002;23: 847– 849. 52. Laine FJ, Braun IF, Jensen ME, Nadel L, Som PM. Perineural tumor extension through the foramen ovale: evaluation with MR imaging. Radiology 1990;174:65–71. 53. Kaylie DM, Wax MK, Weissman JL. Preoperative facial muscle imaging predicts facial function after facial nerve grafting. AJNR Am J Neuroradiol 2003;24:326 –330. 54. Schmalfuss IM, Tart RP, Mukherji SK, Mancuso AA. Perineural tumor spread along the auriculotemporal nerve. AJNR Am J Neuroradiol 2002;23:303–311. 55. Chang PC, Fischbein NJ, McCalmont TH, et al. Perineural spread of malignant melanoma of the head and neck: clinical and imaging features. AJNR Am J Neuroradiol 2004;25:5–11. 56. Babbel RW, Harnsberger HR. A contemporary look at the imaging issues of sinusitis: sinonasal anatomy, physiology, and computed tomography techniques. Semin Ultrasound CT MR 1991;12:526 – 540. 57. Weissman JL, Tabor EK, Curtin HD. Magnetic resonance imaging of the paranasal sinuses. Top Magn Reson Imaging 1990;2: 27–38. 58. Som PM, Dillon WP, Fullerton GD, Zimmerman RA, Rajagopalan B, Marom Z. Chronically obstructed sinonasal secretions: observations on T1 and T2 shortening. Radiology 1989; 172:515–520. 59. Williams 3rd DW. An imager’s guide to normal neck anatomy. Semin Ultrasound CT MR 1997;18:157–181. 60. Sigal R. Oral cavity, oropharynx, and salivary glands. Neuroimaging Clin N Am 1996;6:379 – 400. 61. Kurabayashi T, Ida M, Yasumoto M, et al. MRI of ranulas. Neuroradiology 2000;42:917–922. 62. Kahn JL, Wolfram-Gabel R, Bourjat P. Anatomy and imaging of the deep fat of the face. Clin Anat 2000;13:373–382. 63. Braun IF, Hoffman Jr JC. Computed tomography of the buccomasseteric region: 1. Anatomy. AJNR Am J Neuroradiol 1984;5: 605– 610. 64. Braun IF, Hoffman Jr JC, Reede D, Grist W. Computed tomography of the buccomasseteric region: 2. Pathology. AJNR Am J Neuroradiol 1984;5:611– 616. 65. Bielamowicz SA, Storper IS, Jabour BA, Lufkin RB, Hanafee WN. Spaces and triangles of the head and neck. Head Neck 1994;16: 383–388. 66. Ikeda M, Motoori K, Hanazawa T, et al. Warthin tumor of the patorid gland: diagnostic value of MR imaging with histopathologic correlation. AJNR Am J Neuroradiol 2004;25:1256 –1262. 67. Motoori K, Iida Y, Nagai Y, et al. MR imaging of salivary duct carcinoma. AJNR Am J Neuroradiol 2005;26:1201–1206. 68. Takagi Y, Sumi M, Sumi T, Ichikawa Y, Nakamura T. MR microscopy of the parotid glands in patients with Sjo¨gren’s syndrome: quantitative MR diagnostic criteria. AJNR Am J Neuroradiol 2005; 26:1207–1214.

Wippold 69. Patel RR, Carlos RC, Midia M, Mukherji SK. Apparent diffusion coefficient mapping of the normal parotid gland and parotid involvement in patients with systemic connective tissue disorders. AJNR Am J Neuroradiol 2004;25:16 –20. 70. Kalinowski M, Heverhagen JT, Rehberg E, Klose KJ, Wagner H-J. Comparative study of MR sialography and digital subtraction sialography for benign salivary gland disorders. AJNR Am J Neuroradiol 2002;23:1485–1492. 71. Curtin HD. Separation of the masticator space from the parapharyngeal space. Radiology 1987;163:195–204. 72. Fruin ME, Smoker WR, Harnsberger HR. The carotid space in the suprahyoid neck. Semin Ultrasound CT MR 1990;11:504 – 519. 73. van den Berg R, Verbist BM, Mertens BJA, van der Mey AGL, van Buchem MA. Head and neck paragangliomas: improved tumor detection using contrast-enhanced 3D time-of-flight MR angiography as compared with fat-suppressed MR imaging techniques. AJNR Am J Neuroradiol 2004;25:863– 870. 74. Tryhus MR, Smoker WR, Harnsberger HR. The normal and diseased masticator space. Semin Ultrasound CT MR 1990;11:476 – 485. 75. Hardin CW, Harnsberger HR, Osborn AG, Doxey GP, Davis RK, Nyberg DA. Infection and tumor of the masticator space: CT evaluation. Radiology 1985;157:413– 417. 76. Rossiter JL, Hendrix RA, Tom LW, Potsic WP. Intramuscular hemangioma of the head and neck. Otolaryngol Head Neck Surg 1993; 108:18 –26. 77. Parker GD, Harnsberger HR, Jacobs JM. The pharyngeal mucosal space. Semin Ultrasound CT MR 1990;11:460 – 475. 78. Lanzieri CF, Bangert B. Magnetic resonance imaging of the nasopharynx. Top Magn Reson Imaging 1990;2:39 – 47. 79. Bolzoni A, Cappiello J, Piazza C, et al. Diagnostic accuracy of magnetic resonance imaging in the assessment of mandibular involvement in oral-oropharyngeal squamous cell carcinoma: a prospective study. Arch Otolaryngol Head Neck Surg 2004;130: 837– 843. 80. Imaizumi A, Yoshino N, Yamada I, et al. A potential pitfall of MR imaging for assessing mandibular invasion of squamous cell carcinoma in the oral cavity. AJNR Am J Neuroradiol 2006;27:114 – 122. 81. Gordon AR, Loevner LA, Shukla-Dave A, et al. Intraobserver variability in the MR determination of tumor volume in squamous cell carcinoma of the pharynx. AJNR Am J Neuroradiol 2004;25:1092– 1098. 82. Davis WL, Harnsberger HR, Smoker WR, Watanabe AS. Retropharyngeal space: evaluation of normal anatomy and diseases with CT and MR imaging. Radiology 1990;174:59 – 64. 83. Davis WL, Smoker WR, Harnsberger HR. The normal and diseased retropharyngeal and prevertebral spaces. Semin Ultrasound CT MR 1990;11:520 –533. 84. Hasegawa Y, Matsuura H. Retropharyngeal node dissection in cancer of the oropharynx and hypopharynx. Head Neck 1994;16: 173–180. 85. Smoker WR, Harnsberger HR. Differential diagnosis of head and neck lesions based on their space of origin. 2. The infrahyoid portion of the neck. AJR Am J Roentgenol 1991;157:155– 159. 86. Casselman JW, Biebau G. Imaging of laryngeal cancer. Acta Otorhinolaryngol Belg 1992;46:161–174. 87. Vokes EE, Weichselbaum RR, Lippman SM, Hong WK. Head and neck cancer. N Engl J Med 1993;328:184 –194. 88. Giron J, Joffre P, Serres-Cousine O, Bonafe A, Senac JP. Pretherapeutic evaluation of laryngeal carcinomas using computed tomography and magnetic resonance imaging. Isr J Med Sci 1992; 28:225–232. 89. Sakai F, Sone S, Kiyono K, et al. MR evaluation of laryngohypopharyngeal cancer: value of gadopentetate dimeglumine enhancement. AJNR Am J Neuroradiol 1993;14:1059 –1069. 90. Castelijns JA, Gerritsen GJ, Kaiser MC, et al. MRI of normal or cancerous laryngeal cartilages: histopathologic correlation. Laryngoscope 1987;97:1085–1093. 91. Rouvier H. Anatomy of the human lymphatic system. Ann Arbor, MI: Edwards Brothers; 1938. p 1– 82. 92. Kaji AV, Mohuchy T, Swartz JD. Imaging of cervical lymphadenopathy. Semin Ultrasound CT MR 1997;18:220 –249. 93. Som PM. Lymph nodes of the neck. Radiology 1987;165:593– 600.

CT and MRI in Head and Neck Imaging 94. Dillon WP, Harnsberger HR. The impact of radiologic imaging on staging of cancer of the head and neck. Semin Oncol 1991;18:64 –79. 95. Maeda M, Kato H, Sakuma H, Maier SE, Takeda K. Usefulness of the apparent diffusion coefficient in line scan diffusion-weighted imaging for distinguishing between squamous cell carcinomas and malignant lymphomas of the head and neck. AJNR Am J Neuroradiol 2005;26:1186 –1192. 96. Fischbein NJ, Noworolski SM, Henry RG, Kaplan MJ, Dillon WP, Nelson SJ. Assessment of metastatic cervical adenopathy using dynamic contrast-enhanced MR imaging. AJNR Am J Neuroradiol 2003;24:301–311. 97. Sumi M, Sakihama N, Sumi T, et al. Discrimination of metastatic cervical lymph nodes with diffusion-weighted MR imaging in pa-

465 tients with head and neck cancer. AJNR Am J Neuroradiol 2003; 24:1627–1634. 98. Takashima S, Noguchi Y, Takeuchi N, et al. Head and neck carcinoma: detection of extraorgan spread with MR imaging and CT. Eur J Radiol 1992;14:228 –234. 99. Anzai Y, Blackwell KE, Hirschowitz SL, et al. Initial clinical experience with dextran-coated superparamagnetic iron oxide for detection of lymph node metastases in patients with head and neck cancer. Radiology 1994;192:709 –715. 100. Anzai Y, McLachlan S, Morris M, Saxton R, Lufkin RB. Dextrancoated superparamagnetic iron oxide, an MR contrast agent for assessing lymph nodes in the head and neck. AJNR Am J Neuroradiol 1994;15:87–94.